Colorectal cancer (CRC) screening related reductions in incidence and mortality are not benefiting all population subgroups. In clinical settings, repeat attempts at promoting screening are often required over time with the same patients. Few studies have evaluated whether repeated use of interventions in underserved groups can produce greater screening. Multi 'dose' interventions may be necessary to elevate screening rates to levels seen with breast and cervical cancer screening. This study will utilize a Bayesian adaptive design to study CRC screening in a highly diverse, safety-net clinic population. Although a growing number of studies have found implementation intentions-based interventions effective, no studies have tested such techniques twice in sequence to enhance CRC screening uptake. CRC screening implementation intentions address the when, where and how details of stool sample or colonoscopy screening. A randomized design will test a comparison condition of generic health education (HE) versus implementation intentions (I2), and, for participants who do not complete screening within six months, will re-randomize to one of these two strategies again. All intervention materials will be delivered via low-cost touch screen computers and tablets customized for use in English, Spanish, Nepali, Burmese, and Arabic for Somalis (to accommodate safety-net refugee clients). Pilot testing and formative phase focus groups will test the salience and appropriateness of touch screen audio, video, and graphic materials for each of these groups (except Nepali, where materials were recently developed and pilot tested). The study will be conducted with 500 patients eligible for CRC screening in a group of 9 safety- net primary care clinics. All participants will complete a touch screen administered screening form, informed consent, a baseline survey, and receive information on test options. Because of the rapidly changing insurance landscape for the underserved, we will not cover test costs beyond coverage already in place or obtained for each participant. Participants who complete screening within 6 months will only receive HE or I2 once. Those who do not screen will receive a second intervention dose and re-assessment of insurance status to test varied combination effects of HE +HE, HE + I2, I2 + I2, or I2 + HE on CRC screening completion. Changes or variations in health coverage will be analyzed as a co-variate. Six and 12-month phone surveys will be completed with all participants. The primary outcome will be CRC screening completion at 12 months. Secondary outcomes will come from a mediation analysis of the main outcome and co-variates. We will conduct a cost analysis of the intervention and its variable components. This study will provide important information on the utility and feasibility of embedding low-cost, implementation intentions-based interventions in real-world primary care, where they may be cost effective for repeated application over time with difficult to promote behaviors such as CRC screening.